Clinical specimens and TCGA data analysis
89 pairs of HCC tissues and peritumor samples for quantitative real-time polymerase chain reaction (qRT-PCR) analyses, 10 pairs of HCC tissues and peritumor samples for western blot analyses and 4 pairs of paraffin-embedded tissue for immunohistochemistry (IHC) analyses were obtained from patients at Zhongshan Hospital of Fudan University (Shanghai, China) from 2004 to 2005. Ethical approval was examined and certified by the Ethics Committee of Zhongshan Hospital Biomedical Research Department, and written informed consent was obtained from all involved patients. The slides of paraffin-embedded HCC tissues and paired peritumor samples for IHC analyses were purchased from Liaoding (Shanghai, China) (n=20). The whole transcriptome sequencing (RNA-seq) data of 374 liver tumor tissues and 50 adjacent non-tumor tissues were obtained from The Cancer Genome Atlas (TCGA) liver cancer dataset (LIHC) (http://cancergenome.nih.gov). mRNA level of COL4A1 of 20 tumor types and another 4 datasets of HCC were obtained from the Oncomine database (https://www.oncomine.org/resource/main.html). Representative IHC staining results of COL4A1 in HCC and normal liver tissues were obtained from the Human Protein Atlas online database (https://www.proteinatlas.org, magnification, x40).
Cell lines and culture
Human HCC cell lines (HepG2, PLC/PRF/5, Hep3B, and SK-Hep1) were purchased from the American Type Culture Collection (Manassas, VA, USA). HCC cell line (Huh7) was provided by Riken Cell Bank (Tsukuba, Japan). Human normal liver cell line (L02) and HCC cell line (SMMC7721) were purchased from Cell Bank of the Chinese Academy of Sciences (Shanghai, China). All cell lines were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Gibco, USA) supplemented with 10% fetal bovine serum (FBS; Gibco, USA), 100 units/ml penicillin and 100 μg/ml streptomycin (Gibco, USA), and incubated at 37°C in a humidified incubator containing 5% CO2.
Antibodies and plasmids
Antibodies used in this study were shown in Additional file 1: Table S1. Human expression vectors for HA-RUNX1, Myc-RUNX2 and Myc-RUNX3 recombinant proteins were generated using pCMV backbone vector. The promoter of COL4A1 was cloned into pGL3 basic vector. The subcloning primers were listed in Additional file 2: Table S2.
Western blot
Cells were washed three times with cold phosphate buffered saline (PBS) and total cellular protein was extracted using RIPA lysis buffer (Qiagen, Germany) supplied with proteinase inhibitor cocktail and phosphatase inhibitor (Roche Applied Science, Switzerland). The lysates were incubated on ice for 30 min followed by centrifugation at 4°C 12000xg for 30 min. Protein concentrations were analyzed using the Bicinchoninic Acid (BCA) Kit (Pierce, Rockford, IL). 40 μg of total proteins were separated by 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto 0.22 μm polyvinylidene fluoride membrane (PDVF; Millipore). The membranes were blocked with 5% non-fat dried milk for an hour at room temperature, and then incubated with primary antibodies overnight at 4°C. Protein bands were visualized by the enhanced chemiluminescence (ECL) detection kit (Tanon, China). Quantification of western blots were analyzed by Image J.
Immunohistochemistry (IHC)
Paraffin-embedded HCC tissues and peritumor samples were completely deparaffinized, and then performed antigen retrieval using antigen retrieval solution (Beyotime, Shanghai, China). Immunohistochemistry staining were applied with Immunohistochemistry Application Solutions Kit (Cell Signaling Technology, Dancers, USA) according to the manufacturer’s instructions. Briefly, the slides were incubated with primary anti-COL4A1 (1:200, Abcam) overnight at 4°C and incubated the secondary antibody 30 min at room temperature. Then, the slides were counter stained with hematoxylin for 3 min. Staining results were independently evaluated by two experienced pathologists who were blinded to all clinical data.
RNA extraction and qRT-PCR
Total RNA was extracted from cell lines and tissue samples using the TRIzol kit (Invitrogen, Carlsbad, CA, USA). RNA (1 μg) was reverse-transcribed into cDNA immediately using Prime-Script RT kit (Takara, Shiga, Japan) following manufacturer’s instructions. qRT-PCR was carried out with SYBR Premix EX Tag (Takara) on an ABI Prism 7500 fast RT-PCR instrument (Applied Biosystems, Foster City, CA). Each experiment was performed in triplicate. β-actin was used as the internal reference gene. Data were acquired during the extension step. Objective CT values were normalized to β-actin and 2-△Ct method was used to calculate relative mRNA levels of gene expression. Primer sequences for qRT-PCR were listed in Additional file 2: Table S2. The qRT-PCR primers for COL4A1 are not exon spanning type, but their specificity has been tested.
Lentiviral constructs and cell infection
To knockdown COL4A1 in cell lines, two independent shRNA sequences were designed and cloned into the pGreen-Puro vector (System Biosciences, CA). Another shRNA with a non-targeting sequence was used as a negative control (NC). The shRNA sequences were listed in Additional file 2: Table S2. Virus packaging was performed in HEK 293T cells after co-transfection of pGreenPuro-shCOL4A1 with packaging plasmid pPACK-GAG, pPACK-REV (System Biosciences) and envelope plasmid pVSV-G (System Biosciences) using Lipofectamine 3000 (Invitrogen, Carlsbad, CA, USA). Viruses were harvested 48 h after transfection. Medium with viral supernatant was filtered through a 0.45 μm strainer and viral titers were determined. SMMC7721 cells and SK-Hep1 cells were infected with lentivirus using polybrene (6 μg/ml, Sigma).
CRISPR/Cas9 Synergistic Activation Mediator (SAM) is an engineered protein complex, which is a powerful tool for strong transcriptional activation of endogenous genes at targeted sites. Complete SAM system consists of two separate lentiviral vectors: the dCas9-VP64-puro vector and the sgRNA (COL4A1)-MS2-P65-HSF1-G418 (single guiding RNA) vector. Plasmids design and two lentiviruses packaging were done at Genechem Co., Ltd. (Shanghai, China). sgRNAs with matching COL4A1 gene promoter sequences (Gene Bank ID: NM_001845) were listed in Additional file 2: Table S2. HepG2 cells and PLC/PRF/5 cells were infected with the lentivirus of dCas9-VP64-puro. After two weeks of puromycin (1 μg/ml) selection, cells were infected with indicated sgRNA (COL4A1)-MS2-P65-HSF1-G418 lentivirus and selected with G418 (600 μg/ml).
siRNA knockdown
The siRNAs targeting human COL4A2, COL3A1, COL1A1, RUNX1 and the scramble siRNA control were obtained from Biotend (Shanghai, China). siRNAs were transfected into indicated cells with Lipofectamine 3000 according to manufacturer’s instructions. Cells were harvested 48-72 hours post transfection for various assays. Sequences for siRNA were listed in Additional file 2: Table S2.
Proliferation, migration and invasion analysis
Cell proliferation, invasion and migration assays were measured with the xCELLigence System’s Real-Time Cell Analyzer (RTCA, Roche/ACEA Biosciences) placed in a humidified incubator and maintained at 37°C with 95% air/5% CO2. This system continuously monitored electrical impedance which created by cell adhesion and proliferation in microelectrode-integrated membrane, and outputted as a unit-less parameter (cell index). For proliferation assays, 1× 104 to 3× 104 cells were seeded into E-plate 16 (ACEA Biosciences) with 200 μL DMEM containing 10% FBS (n=3). Cell index was normalized to baseline reading at time point 0, and measured every 30 min for 72 h. Migration and invasion assays were performed in 16-well CIM plates (ACEA Biosciences). For migration assays, 1.5× 105 cells were seeded as triplicates in the upper chamber in serum free medium. Upper chamber was then placed on the lower part of the CIM-device containing DMEM with 10% FBS as a chemoattractant. Cell index was measured every 30 min for 48 h. For invasion assays, upper chamber of CIM-16 plate was initially coated with Matrigel (BD Biosciences, Bedford, MA, USA) diluted in serum free medium at a ratio of 1:20. Then, next steps were same with migration assays.
For cell proliferation assay, Cell Counting Kit-8 (Dojindo, Japan) was also applied according to the manufacturer’s instructions. Briefly, 3× 103 cells per well were planted in a 96-well plate. Absorbance at OD450 was measured for 4 consecutive days and used to plot cell growth curves. For colony formation assay, 8× 102 cells were seeded in 6-well plates and maintained in DMEM medium with 5% FBS. After 14 days, cells were washed with PBS and stained with 0.5% crystal violet. For transwell assay, 5× 104 cells were seeded into upper chambers (transwells with 8-μm pores, Corning, USA) without FBS, and DMEM containing 10% FBS was introduced to the lower chambers (24-well plates). After 2 days, cells remaining on the upper surface of the filter were removed using a cotton swab with PBS, then transwells were fixed and stained with 0.5% crystal violet. The migratory cells were photographed and counted in 5 different fields per well.
Drug treatment
FAK inhibitor Defactinib (VS-6063) was purchased from CSNpharm (CSN17445, Shanghai, China). Src inhibitor Saracatinib (AZD0530) was purchased from Selleck Chemical (S1006, Huston, TX, USA). To evaluate inhibitory activity of FAK or Src inhibitor (Defactinib or Saracatinib), cells were firstly seeded at a density of 5× 103 in 96-well plates and incubated overnight. Then Defactinib or Saracatinib was added at indicated concentrations. After 2 days, CCK8 was applied to measure survival cells following manufacturer instructions.
Wound healing assay
Cells were grown in 12-well plates at 95% confluency. A linear wound was scratched with a 200 μL sterile pipette tip across the monolayers. After washing with PBS to remove cell debris, adherent cells were incubated in medium with 10% FBS. Wounded monolayers were photographed every 3 h for 24 h.
Tumor xenograft models
Subcutaneous xenograft mouse model was used to assess tumor growth. Animal experiments were approved by the Ethics Committee of the Renji Hospital, Shanghai Jiao Tong University School of Medicine. Female nude mice (age, 4–5 weeks; weight, 14–16g; Institute of Zoology, Chinese Academy of Sciences) were randomly divided into three groups: two COL4A1 knockdown groups and one NC group (n = 7 per group). A total of 2 × 106 SMMC7721 cells in 100 μL of DMEM without FBS were injected into right axillary fossa of nude mice. Tumor volume was measured by caliper measurements every 3 days and calculated with the formula of (length × width^2)/2.
Statistical analysis
Data were analyzed using GraphPad Prism 7. Results were presented as mean ± standard deviation (SD, n=3). Statistical differences between groups were evaluated by the student’s t test (paired/unpaired). Pearson correlation tests were performed on correlation analyses. Two-way analysis of variance (ANOVA) followed by Tukey’s multiple comparisons test was performed to compare significant difference and calculate the P-value between the different groups. It was considered as statistically different when P < 0.05 (*P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001), otherwise not significant (ns).